Existing techniques for generating multi-image scenes with a camera, such as panoramic pictures constructed from multiple images, are typically manual in nature, and usually require trial and error to generate the desired results. One such technique involves placing a camera on a tripod, and placing a protractor under the camera. A first picture is taken, the camera is rotated a fixed number of degrees, and then a second picture is taken. This process may be repeated to take any desired number of pictures. For a conventional analog camera, the film is then developed, and an individual determines from the pictures whether there is any undesired overlap or spacing between the pictures. If there is any undesired overlap or spacing, additional pictures of the scene may be taken, using a different amount of rotation than was used previously. Thus, trial and error is needed to obtain the optimum amount of rotation of the camera to generate images for a multi-image scene.
The above-described prior art technique may also be applied to a digital camera. In addition, some existing digital cameras include a built-in display that allows a user to attempt to “line-up” currently viewed images with a previously captured image, to facilitate capturing images for a multi-image scene. For example, a previously captured image or portion of an image is displayed on one half of the built-in display, and a currently viewed image, or portion of an image is displayed on the other half of the built-in display, and a user manually snaps a picture when the images appear to line-up. This process relies on the subjective judgment of the individual to line-up successive images.
It would be desirable to provide a camera that automatically determines appropriate times to capture images for a multi-image scene, without requiring the user to manually line-up successive images.